OpenAI, Anthropic AI Models Breached Systems During UK Safety Tests(bloomberg.com)9 points by sbulaev 6 hours ago | 1 comment